17February
Iain Dale reports of a plot to oust Patrick Cormack. This would be a bad idea, for Mr. Cormack has been a remarkably good constituency MP. The Snow in the Summer was brought up in the heart of South Staffordshire, and was represented by this gentleman for over a quarter of a century. Whenever we heard him, Mr. Cormack was polite, tolerant of different views, and quietly persuasive. Whenever we wrote to him, we got the impression that he was going to do whatever he could to resolve the matter, or listen to our views. Richard Burden, the current MP for this blog, may be good, but Patrick Cormack is the master at nursing a constituency. Heck, he's been in parliament for so long that the one-nation Conservatism he espouses has been fashionable, unfashionable, and is now undergoing a bit of a resurgence.
Our previous post on the swingometer crossed in the blogosphere with Anthony Wells's take on the situation. He reckons that the Conservatives aren't doing enough in the national opinion polls to win an overall majority next time out; our analysis suggests they might, just, scrape that overall majority. If we were betting people, we would be piling into No Overall Majority for UK-2010, but with a small saver on a small (20 or under) Conservative majority.
The swingometer model includes an element of anti-Labour tactical voting, as follows:
* Labour-held seats, where Labour's lead is no more than 20% and the second-placed party is at least 5% clear of the third;
* Opposition-held seats, where Labour is in second place, within 10% of the lead, and no other party is within 10%;
In these seats, one-seventh of the Labour vote will divert to the opposition challenger or incumbent. One voter in seven was the anti-Labour tactical vote at the 2005 election, and in the absence of research into this phenomenon, we propose to leave this part of the model static.
The model also incorporates a reversion factor, which currently assumes that for every 19 voters transferring between parties, 10 will return at the next election. The 19 reduces by one per year, at the anniversary of the election, so even at the end of a five-year parliament, only two-thirds of the transfers will be retained.
The model does assume a Uniform National Swing, simply because there is too little evidence to generalise further. In the three months to this week, there have been just 32 by-elections across the UK producing a usable swing; trying to split that further would leave the model liable to inaccuracies of If the whole of the North behaved like Lancaster Riverside.... However, the UNS is applied by proportion of the vote, rather than a straight addition and subtraction. For instance, Labour is currently releasing 10.3% of the nation's voters, 28% of its vote at the last election. I'm reducing Labour's vote by 28% in each constituency, then allocating those voters in the ratio 5 Conservatives to 4 Lib Dems to 1 Other, and then adjusting for tactical voting. We understand that this proportionate model is the same one that the BBC and ITN use, but most amateur psephologists use the much simpler additive model.
If reflected at a general election, and updating for this week's results (slightly better for Labour than the last few months) the probability distribution of the results is as follows:
Conservatives the largest party - almost certain (greater than 99%).
Conservatives have an overall majority - about 65%.
Conservative overall majority of 10 - single most likely result.
Conservative overall majority of 20 - about 30%.
Conservative overall majority of 44 - about 5%.
Those are the facts. Here is opinion.
The election will be decided in the marginals, as ever. Labour's fate will be sealed in the Midlands, in London, in the urban centres of the North, in the Celtic fringe. The Conservatives' fate will be decided in their marginals, in the South, in the Midlands, in rural parts of the North and Scotland. Most of these seats are marginals against the Lib Dems, not against Labour.
The psephology makes it particularly difficult to engineer a hung parliament. Squeeze Labour too hard in the cities and the bandwagon ensures that the Lib Dem rural marginals turn blue; leave the rural marginals alone and Labour won't lose enough seats to drop below the Tories.
The wild card, just as it was in 1974, is the possible growth of the Nationalist parties. There are 6 Scottish Nationalist members, and 3 Plaid Cymru MPs. It is possible that Plaid can pick up one or two gains, but the Scottish indepence movement seems to be gathering steam rapidly. All bets are off if the SNP is the largest party at this May's elections, and is subsequently denied a place in government; they could start taking protest votes against parties deemed responsible for perverting democracy.
The other matter of interest is the fate of the independents. George Galloway has said that he'll be stepping down from Bethnal Green, so chalk that up as a Lab-LD marginal. The strange situation in Merthyr Tydfil cannot last forever, and the seat will remain an Ind-Lab marginal. Wyre Forest, held for the last two parliaments by the Health Campaign, is now an Ind-Con marginal, perhaps liable to go Con following boundary changes. Might Patrick Cormack in Staffordshire S become the first independent Conservative elected since Mr. Robertson in Caithness and Sutherland, back in 1959? He didn't hold the balance of power; on current form, an independent Mr. Cormack might.
Is there a masterplan for delivering a tightly-hung parliament? No; a few hundred votes in a dozen seats could tip the balance in any direction. We will watch the Scottish election closely, and return to the fray much later on.
